As I recall from Arnold Bank, my old calligraphy teacher, the ct and st ligs were originally done in calligraphy. I think by dragging the corner of the flat pen or using low pressure on the pointed pen (and Wacom thought they started pressure sensitive drawing:-) Look at Blackletter writing as well as cursive. Those old guys were really good at it.

But what order are the parts of "c" and "s" [typically] rendered with a broad-bin pen - top then bottom(s)? And in ballpoint pens it certainly goes top-down, no? So this is why the connection on top seems strange to me. Intentional "design" so to speak - not "writing-facilitory". No?

hhp

I don't remember the stroke order or even if this was done the same by most writers. It was more on the order of a flourish and sometimes a way to make up for a spacing dilema. An example is the end-of-line "e" which had a long extension below the crossbar. Remember, letter and word spacing was done by eye and on the fly; sometimes the end of a line came unexpectedly so little tricks and alternate letterforms came about to save the day. Later these variants became normal and sometimes could be used to distinguish individual hands.

Interesting.
So I'm guessing most ligatures were towards the right side of a column... :-)
So maybe that's another way for somebody to make typography look more handwritten: do funnier stuff the closer the linebreak.

hhp

The 'ct' and 'st' ligatures have nothing to do with the line break; they can occur anywhere in the line. There's a little more functional sense to it than that.

One way of writing a 'c' is in *two* strokes:

Start at the thinnest part of the c (about 10:00) and continue down counterclockwise until the end of the lower arm.

Then lift and go back to the starting point of your first stroke, but now proceed clockwise to finish the top arm of the c.

If a 't' follows, you are now in a good position to loop upwards to the top of the stroke.

The same applies, mutatis mutandis, to an 's'.


Victor

OK, interesting. I figured the "classical" construction is in multiple strokes, but didn't realized you'd start at the bottom. But since the "origin" is above-center, it makes more sense now. Thanks.

So chiro it is (at least mostly).

hhp

Hrant, I'm late to the party, but I have to agree that it seems the origin is definitely chirographic. I made some illustrations for what they might be worth.

A browse through some calligraphy instruction manuals show pretty consistent construction (I'm not a calligrapher but the books had letters in them...):



Here's a detail from an illustration in Fairbank's A Handwriting Manual:



The highlight box shows typical construction of ss, ct, and st ligatures. Of note, but not highlighted is the flourished st ligature above the highlight box in the word 'est'.

Fairbanks asserts that 'Joins are undoubtedly agents of speed'. This seems logical in the italic hand but here is an example of the ct and ss ligatures in use in a book hand (Fairbanks, Plate 2 detail):



I guess it would preserve speed here too, though it looks as though there is no pretense at flourish or decoration here.

I noticed (but did not scan) the text in Morison's Letter Forms which is set in Digital Fournier. The ct and st ligatures have lost most (not all) of their chirographic thick and thin and seem to be more sculpted into the serif on the c and s and into the top of the t (as contrasted with the lower left highlight in the last Psalter example).

...sometimes the end of a line came unexpectedly so little tricks and alternate letterforms came about to save the day

Which is true and many letters that don't have crossbars end up with crossbars or a horizontal exit bar when this happens like the lc n occuring at the end of a line that needs justification to the one preceding.

I hope any of this is useful.

Jon

Wow, nice treatment - definitely useful - thanks.
So the top-last scheme in the "c" and "s" seems pretty standard. Although in the Camp method, instead of a loop to the top of the stem I guess you'd just get a continuation into the bar - which however I have never noticed happen.

The Mercator ligation seems a bit mannered, no? It makes me think that, yes, the original motivation was chirographic facilitation, but then formal sensibilities kicked in. This, even though you're right in alluding to the fact that slanted/cursive writing is faster/less_formal. But the Psalter sample is much more sobre, and it still have ligation, albeit logically more reserved instances.

> Morison's Letter Forms .... more sculpted

And to me those thin joins are awkward and distracting; sort of like the "perfect"-circle OS zero.

--

All this makes me realize that "discretionary" isn't a really good term. For one thing, originally it was apparently an issue of facilitation (not caprice), plus it makes me ask: is the "fi" ligature in fact more important than the "ct" and "st", or is it perhaps that an "f" with a strong beak is in fact itself "discretionary"?!

hhp

Wow, nice treatment - definitely useful - thanks.
My pleasure.

Although in the Camp method, instead of a loop to the top of the stem I guess you'd just get a continuation into the bar - which however I have never noticed happen.
Yes you definitely would. The examples that she shows are consistent with this. Her style also doesn't ligate except in tt, ra (which looks cool) and the ear on the g lending itself as the dot on i in a gi combination.

The Mercator ligation does seem mannered especially in the word est and in the ct, though I think in the st ligation using the long s makes it look more natural.

I think that the thin joins in Digital Fournier are definitely distracting. I asked my lady to read a page of the text and tell me what she thought, and she said those connectors on the s and t make me crazy. So in my informal survey, we're not the only ones.

I have to say that ct and st ligatures can sure look cool. I like the idea of them. I think that the space between the letters having an arcing top could be nice like in the psalter example. I'd definitely consider including these ligatures in an italic companion to a text face.

Personally, I think that ff fi tt are more important.

As for the 'discretionary' descriptor, the employement of any ligatures would seem to be at the discretion of the typesetter, as their design is at the discretion of the type designer. Not really wrong but...

It may be more useful to make a distinction about how critical a ligature is. In this regard I would say that fi is more critical than ct, as fi affords greater opportunity for letter shape collision and murkyness than st. Also since the st is a chirographic descendant, I would argue that it's existence is less critical and more appropriate to italic forms (though not exclusively). I think I would enjoy seeing a text face where the ct and st looping isn't as sculptered as in Digital Fournier, but more integrated as in the psalter example. And I could imagine that it could look very attractive...

Jon

>If a 't' follows, you are now in a good position to loop upwards to the top of the stroke.

Sorry Victor, the writer is in a difficult position.
To make the typical ct or st join according to the technique you describe, one has to push upwards (bottom right to top left) from the end of the c or s, digging the pen into the paper.

However, it IS possible to do such a "backwards" stroke, keeping the pen barely touching the paper, and the stroke brief.

And it's a skill worth mastering, because the alternative is to bring one's pen back to the same spot three times, and each time to halt and raise/drop the pen: once at the end of the final "c" stroke, once at the beginning of the "t " diagobal upstroke (see Camp, in above JW post), and once again for the beginning of the "t" crossbar. This does not feel like good, flowing calligraphy, and I think this is one reason why calligraphy has developed the "ct" and "st" ligatures, to avoid this constricted series of motions.

Another chirographic reason: from an aesthetic point of view, the combinations ct and st are unsatisfying, because the horizontal strokes at the top of the "x height" are close, and yet one cannot render them as one horizontal joining stroke (as in ti and fi), because that would mess up the c/s (for the same reason that one does not top-join ci and si as ligatures). So, finishing the c/s with a flourished ligature provides a sense that the character has been concluded purposefully.

Why did the ct and st ligatures persist in type when so many others were dropped?

Because the humanist t with its very short ascender is a weak glyph. It looks quite similar to r and c in old-style faces from Bembo to Caslon. The ct and st ligatures add distinction, in more ways than one.

To make the typical ct or st join according to the technique you describe, one has to push upwards (bottom right to top left) from the end of the c or s, digging the pen into the paper.

It is worth remembering that in the Renaissance reed pens were in more widespread use in Europe than they were in later times, when they were almost wholly supplanted by steel nibs. As any Arabic calligrapher can ably demonstrate, a reed pen can be pushed without 'digging into the paper'. In addition to widespread use of reed pens, contemporary instructions for the cutting of quills stress suppleness as a desirable quality.

I know, no one wants my input?

Renaissance reed pens

This would make a lot of sense. Especially for the { ct }

It would appear, at first glance, most certainly in Caslon's time, {famous for the long s} that the {st } tied letters

>'digging into the paper'.

John, digging in to the paper is the worst case scenario.
A more likely event is having one's nib "catch" on a piece of fibre.

A supple pen (especially with a fine point) can be a disadvantage when doing a push stroke, because it will start to bend and catch at the slightest resistance, whereas a stiffer (and/or broader) nib will push through, or ride over, resisting fibres and hold its angle.

But suppleness is not a simple issue.

The structure of a nib, whether in a reed, quill, or metal, is designed to be supple when pushed down, yet stiff in other directions -- not only does this allow the user to regulate ink flow by the amount of downward pressure, it also keeps the nib firm when moved sideways or pushed, when stiffness is required to prevent the nib from catching.

The relationship between the hardness of the nib material and the smoothness/hardness of the paper/vellum is also an issue.

Whatever the pen, going against the grain is counter-intuitive to the way one has learned to write calligraphy, and the way the vast majority of characters are formed.

However, strokes that are difficult for a nibbed instrument are no problem for a pencil or crayon, and calligraphers develop their embellishments -- complex flourishes -- as extemporized single strokes. So there is definitely a reason to incorporate the more difficult strokes into one's style.

Interestingly, a determining factor as to whether one executes a difficult ligature in a single move involving pushes, or splits it to avoid them, is how much ink is left in one's pen, and how smoothly it's flowing. If one has recently dipped, it's easier to do push strokes.

"So, doesn't this look better than Times?"
"Yeah, I guess so. But what's with all these curly things on the s-t?"
"Those are ligatures."
"I don't like them."
"Why not?"
"I don't know, they look kinda pretentious."
"Pretentious? Why?"
"I don't know, they just do."
"Dammit, Ma."

"So, doesn't this look better than Times?"
"Yeah, I guess so. But what's with all these curly things on the s-t?"
"Those are ligatures."
"I don't like them."
"Why not?"
"I don't know, they look kinda gay."
"GAY?! WTF?!??! Why?!"
"I don't know, they just do."
"Dammit, Dad."

STATE OF FLORIDA BANS LIGATURES IN ALL OFFICIAL DOCUMENTS

SF CHRONICLE ADOPTS LIGATURES IN BODY COPY IN FLORIDA PROTEST

Dear Editors:

Please cancel my subscription to your newspaper. After your overly trumpeted transition to your newfangled typesetting machines, I find I can no longer read your body copy without feeling irritated by these new, howyousay, "ligatures."

Nobody seems to be able to convince me what purpose they serve, and I find that instead of getting annoyed by the President like I should, I keep get annoyed at the damn font. Otherwise I suppose it's nicer than the last. Or at least I take your word for it. I wouldn't know how to measure that, though.

And I'm not convinced these "ligature" things have anything to do with gay people. Perhaps you don't either and are simply mocking the notion. But this has gone on for two weeks now and the joke isn't funny anymore. Get over yourselves.

Bitterly Yours,
Frazzled in Fresno

FOCUS ON THE FAMILY DECRIES GROWING USE OF DISCRETIONARY LIGATURES

SENATOR ATTACKS FUNDING OF DISCRETIONARY LIGATURES

VATICAN CLARIFICATION: VERSICLE AND RESPONSE SYMBOLS ARE NOT LIGATURES

FHQWGADS TYPE FOUNDRY ANNOUNCES LIGATURE-FREE VERSIONS OF ALL TYPEFACES

LOL!!!
John,
That is a scream

JB, that's Hall of Fame material.
Right up there with "If God Designed a Typeface".
I think we're about ready for a "Best of Typophile" in print.
If Armin can do it...

A supple pen (especially with a fine point) can be a disadvantage when doing a push stroke, because it will start to bend and catch at the slightest resistance, whereas a stiffer (and/or broader) nib will push through, or ride over, resisting fibres and hold its angle.

Yes, I was probably misleading in mentioning suppleness, rather than the more specific qualities of reed pens (which I've only started using since taking an interest in semitic scripts). Because the outside of the reed is denser and less flexible than the inside, the reed pen will tend to flex when pulled but not when pushed, while the relative softness of the tip (relative to metal nibs) allows it to glide over even quite rough paper. I'm really quite thrilled with my reed pens, although they take some getting used to and managing ink flow is tricky for someone used to the regulated flow of a capilliary nib system.

Whatever the pen, going against the grain is counter-intuitive to the way one has learned to write calligraphy, and the way the vast majority of characters are formed.

Unless you're writing Arabic

I really learned about type by taking a calligraphy course I still have the pens and I'm still asked to ocassionaly to make individiual place cards. Its something about sepia ink with various weights with a persons name that makes it special.

> the ear on the g

Michael Harvey does some amazing ligations with the ears of his "g"s.

Digging/catching: Is it possible that many (most?) "ct" and "st" ligs have thin, monoline connectors because you use the pen differently for them?

John: Very funny!

hhp

If I remember rightly, I read that the st ligature, with the short 's', is not historical. The switch to all short s only came after the French revolution. The long 's' with the t is the historically correct ligature, but here the character extends to the right and it is logical like the fi ligature. Was the ct ligature as common in the old days? Is there also an optical reason to separate the c and crossbar of the t?

William,

Is there also an optical reason to separate the c and crossbar of the t?

No, the ct is a flourish.

>Is there also an optical reason to separate the c and crossbar of the t?

Yes. If the distinction is not sufficient, the crossbar will take over the the top of the c, and it will look like a dotless i. But a flourish is only the most extreme way to make the distinction apparent.

Wasn't it Bell who firft got rid of the long-s?

hhp

William,

Nick says, "Yes. If the distinction is not sufficient, the crossbar will take over the the top of the c, and it will look like a dotless i."

I think Nick

besides "tt" and "ff" my most common ligature in writing is "in" with the dot of the i lending itself to the stem of the n. what other ligatures do you all use in everyday handwriting?

Calvert Guthrie

I think Nick

> setting ligatures never saved any time

This is not true. An entire system (actually, at least two of them) were devised beginning in the second half of the 18th century whereby "logotypes" consisting of frequent sequences of letters -much more than just two- were cast as individual sorts, to save time while composing. The Times newspaper was set in such a scheme in the beginning, but then the printers union -realizing the savings in time, hence reduction in their incomes- decided to charge by the letter and not the sort, so the efficiency of the system became moot. And then mechanical composition came around, so it didn't matter.

And here's the layout from a similar scheme carried out by Henri Didot and Marcellin Legrand:
http://www.themicrofoundry.com/other/polytypie.gif
(Detailing mine.)

hhp

>overwhelming evidence to the contrary

Gerald, pay attention and stop trying to annoy me.
(And please excuse my rusty calligraphy.)

1. shows the problem that can occur between c and t -- the weak ending of the top right of the c, the small gap between it and the crossbar of the t.

2. shows why joining the top of the c with the crossbar of the t is not a good idea -- the form of the "c" is compromised (though not quite so dangerously as you would have us believe).

3. typical form of the ligature.

All you need is a small spot or smudge appearing on the page above Nick's middle example, and you have fait instead of fact.

Having spent many months working with people who study ancient texts, looking at facsimiles of Hebrew Biblical manuscripts, I have developed a healthy respect for spots and smudges on the page and their ability to radically change the meaning of a text. Of course, if your vowel system consists largely of dots, the problems tend to be more numerous.

Plus it looks like "fad", sans smudge!

hhp

Hrant

Bad ideas are frequent, even you can have them.

The Times newspaper was set in such a scheme in the beginning, but then the printers union -realizing the savings in time, hence reduction in their incomes- decided to charge by the letter and not the sort, so the efficiency of the system became moot.

I am sure your meant to say, (so the efficiency of the system became a hoot.)

Boy good thing that did not catch on, Linotypes, Monotypes and "Computers" would have been sold for "doorstops".

I would challenge you over this so-called-union-problem. What is your source? The "Wizard of Ozz?" Not that I disbelieve you. Just that I would like to understand the nature of the objection myself. Unlike yourself I have "a more than considerable hot metal understanding'. I have "a working mans perspective". In other words this system is crapolla.

There were many inventions and logotype schemes. Setting ligatures, (logotype) in hot metal "does not save time". If they did the non union shops would have welcomed them, yes?

The more ligatures, the more memory, the more travel of the hands, arms or body. Remember, Chinese often set type using roller skates. http://lanstontype.com/ChineseWindow.html Maybe the British Unions did not think "roller skating was part of their job description".

The California Job Case has spaces for ligatures, (logotypes). Believe me, they were not located in the major traffic zones. They were kept "out in the back forty". The boxes were so tiny it was hard getting your fingers into them, often the "labour hoot saving logotype ligatures" would jam themselves tight against the sides.

Memory failure is certainly a consideration, if one did not realize at first they were setting a ligature, (logotype) they would have to think, stop, distribute the characters and then go for that obscure box containing "time and labour saving logotype, if only they had known they needed one, remember where it was, and had not loaned their roller skates to the baby sitter".

So don't go all "romantic, poetic and nostalgic on me about hot metal". I can't stand to see a grown digital junky weep.

Get the picture.

Tell me how many ems per hour was set with that silly scheme you have dug up. I can't wait for you to remember all those combinations and where they go. The hoot word breaks would be interesting to watch. Want to make a wager, how about a "race"?

Anyway, that said, I am glad your are interested in this sort of thing. If I ever get to my material I would welcome a more elaborate discussion on the many different inventions. Don't take this as a criticism, it is a backhanded compliment.

> What is your source?

"Printing in the twentieth century" (AKA "The Times book of printing"), 1930. Page 14.

hhp

Hrant,

"Printing in the twentieth century"

Twentieth Century, where

Hrant,

All kidding aside, I am interested in the book. Do you think it is still in print.

If its Monotype the ink is probably not even dry yet. But look, I am interested. Maybe we can do some horsetrading. Sounds like an interesting book.

>you have clearly illustrated why no one would ever have mistake a "c" for an "i"

Have it your way. Others may see things differently.
But that's not the issue.
The purpose of the exercise was not to address the extreme situation of mistaking one letter for another.

I demonstrated, as I intended, that a "c" can indeed look quite like a dotless "i", (and perhaps also create something that looks a bit like a "d") when the "c" and the crossbar of the "t" are not separated. That's all.

The point is not that readers will mistake the letter, but that it is unsatisfactory to have one letter look like another (as with I, l, and 1).

The ct ligature is not a playful ornament that's added on to a perfectly good character combination -- in calligraphy it results in an improved pen motion over doing the letters plain, and it also makes a distinctive glyph to replace a poor character sequence -- where the top terminal of the "c" or "s" meets the crossbar of the "t" indecisively (my left hand illustration, above), neither buffered by a serif, nor smoothly linked (center illo).

Gerald, how do you account for the persistence of the fancy "ct" and "st" serifs, amongst those no-nosense printer boys of yore?

Nick

Gerald, how do you account for the persistence of the fancy "ct" and "st" serifs, amongst those no-nonsense printer boys of yore?

Something I should have addressed originally: those "logotypes" were not ligated, so the reader couldn't tell the difference. So in both cases (Didot and The Times) obviously the only point was to save time, since most humans (there are 1 billion Chinese readers, after all) can handle a lot more than a few dozen glyphs. Sure, a few thousand is physically too much for a composer, but the few hundred in that Didot case seems fine. And note that the more logotypes you have, the smaller the sort boxes need to be, so you save body movements there.

(Not that any of this will make a dent in Gerald's personal religion.)

hhp

Hrant,

Sure, a few thousand is physically too much for a composer, but the few hundred in that Didot case seems fine. And note that the more logotypes you have, the smaller the sort boxes need to be, so you save body movements there.

I get it, you have invented the . . .

"HranTooSoft Keyboard".

Ching ching. Good thinking, I see money in your future. Remember me when you get your yacht.

Question: "Laptops", are they going to be replaced by "Dining Room Tabletops? Or are you going to negotiate a contract with the "Peter Pan Union" for future typesetting workers?

I hope you don't mind me asking.

Nick,

Thanks to you and to John Whipple for the interesting illustrations making clear the calligraphy behind the ligatures. I, at any rate, do appreciate the time you both took to do this. To me the point about the ct ligature is not that you will mistake the ct for this or that, but just that the ligature is an elegant clear solution to a problem, which you point out.

I found on the internet that the short s was used only as a final s, and hence there would be no traditional st ligature in calligraphy. Maybe it is just because I have known about this, but I find the ct ligature nice, in the right setting, but the st ligature consistently annoying. To me the way it redoubles the curve in the s just doesn't look good, and seems affected, unlike the ct ligature.

Am I just imagining this or do some of you folks see the same thing?

> there would be no traditional st ligature in calligraphy.

Since it was a typographer (Bell) and not a calligrapher who initiated the abandonment of the long-s, I'm not sure this makes sense. Presumably calligraphers borrowed this simplification from typography some time down the line, but I'm not sure that's a good cutoff for "traditional" versus not.

hhp

>since it was a typographer (Bell) and not a calligrapher who initiated the abandonment of the long-s,

Actually, that's not so. Calligraphers had for centuries been using their "discretion" over whether to use a long or short s in the middle/beginning of words. (Many examples in "Penmanship of the XVI, XVII and XVIII Centuries, by Lewis F Day).

But was there a "school" of calligraphy that said "we will not use the long-s"?

hhp

William,

Maybe it is just because I have known about this, but I find the ct ligature nice, in the right setting, but the st ligature consistently annoying. To me the way it redoubles the curve in the s just doesn't look good, and seems affected, unlike the ct ligature.

You are a gentleman and a scholar, thank you. I had pointed out that the st was not entirely legitimate. That the "short s" was only used at the end of word therefore could not be followed by a "t".

http://lanstontype.com/Demijohn.html

This specimen was pulled with a "proof planer". The type is set upon the stone, inked, a sheet of paper laid on the surface. Then a block of wood covered in felt is struck with a system of strokes by a wooden mallet. It tends to make the type bold which apparently makes some people really happy.

I was setting a sample of the "long s".

I like the st ligature but not in bookwork. To be honest I don't care if the st ligature is "not historical" because I come from a progressive school of typographers that believe that "new" can be "good". Many in the forum disagree with this premise.

In any event when printing books I choose to loose the "st ligature" but maintain the "ct ligature". Readers are accustomed to the "ct ligature" after 500 years but the "new st ligature is not so acceptable. Besides, I find the frequency rate far too high making them distracting. Also your superior instincts concluding the shape of connector in conjunction with the curve of the s makes it too stressful

Yes and no. In German, the normal s can occur in the middle of a word at the end of a syllable. So "Haustier" would have a nomal s, while "Baustelle" would have a long s. But those same rules dictate that the st in "Haustier" would not ligate. And I've never seen the non-long st ligature in German to begin with.

I'd be curious to know which languages the quaint st and ct appeared set in, and which ones they didn't.

BTW, in English, "ct" is about as common as "fi", and much more common than "fl". And "ffj" of course is pretty useless - even "fb" and "fh" are less useless.

hhp

ct lig practicalelectronicsgernsback.j (178.6 k)

I saw this 1922 magazine cover on http://boingboing.net/ today, and thought this was a more comfortable way of creating a ct ligature, rather than pushing the nib.
Tim

here it is

Tim, that is hand rendered, not written, so yes, no push strokes necessary there.
However, if this form of ligature were written in one pen stroke (except for the crossbar of the "t"), it would involve major "pushing".

Speaking of ligs, how about that "Ov" ?!